Hepatitis E Virus (HEY) was firstly recognized as a pathogen to enterically transmitted non-A, non-B hepatitis in 1983 (Balayan et al., 1983. Intervirology 20:23). Hepatitis E is mainly endemic in developing countries in Asia, Africa and Middle America. In developed countries, hepatitis E cases were mostly found in immigrants or traveler from abroad. Both sporadic cases and large epidemic have been documented. During the period from 1950s to 1990s, several hepatitis E outbreaks happened sequentially due to polluted drinking water (Visvanathan, 1957, Indian J. Med. Res. (Suppl.). 45:1-30; Wong et al., 1980 Lancet., 2:882-885; Myint et al., 1985, Am J Trop Med. Hyg., 34:1183-1189; Belabbes et al., 1985 J Med. Virol., 16:257-263; Hau et al., 1999, Am J Trop Med. Hyg., 60:277-280). Most hepatitis E infection was self-limited and scarcely developed into chronic; but for the pregnant, the sequel was severe with a mortality rate above 17% (Tsega et al., 1992, Clin. Infec Dis., 14:961-965; Dilawari et al., 1994, Indian J. Gastroenterol., 13:44-48; Hussaini et al., 1997, J Viral Hepat., 4:51-54).
In 1991, researchers got the first complete genome sequence of HEV, a single-strand non-enveloped positive RNA virus (Tam et al., 1991, Virology 185:120-131). Sequence analysis showed the genome was 7.2 kb with three open reading frames. ORF1 which locates at 5′ end encodes non-structural protein of the virus, ORF2 which locates at 3′ end encodes major structural protein of the virus. At ORF3 5′ end, there is one by overlapped with ORF1 3′ end. At ORF3 3′ end, there are 339 by overlapped with ORF2 end. It's acknowledged that ORF3 encodes another structure protein with unknown function (Tam et al., 1991, Virology, 185:120-131; Aye et al., 1992, Nucleic Acids Res., 20:3512; Aye et al., 1993, Virus Genes., 7:95-109; Huang et al., 1992, Virology, 191:550-558; Reyes et al., 1993, Arch Virol Suppl., 7:15-25).
The detection of HIV infection mainly depended on Immunological Electron Microscope (IEM) or Immunological Fluorescence Technique for a long time, but those techniques are very complicated, expensive and hard to be fulfilled in many laboratories. After the clone and sequencing of HEV genome, more sensitive techniques like ELISA, Western Blot, PCR, etc. were developed to be used in the detection of HIV infection.
It is well recognized that the development of serum HEV antibody kits is absolutely necessary, but due to very low-concentration HEV virus secreted by the infected human or animals, thus it is impossible to use sera as the source of antigen. Till now, the efficiency of HEV cell culture is still very low, which limited the availability of enough antigens for HEV detection. Thus, the detection of HEV antibody is still depending on synthesized polypeptides or recombinant antigens. Unfortunately, many serological studies showed greatly varied consistence based on synthesized polypeptides or recombinant antigens from different HEV genome regions. For example, Goldsmith et al., (1992, Lancet, 399:328-331) used ORF2 3-2(M) antigen (a.a. 613-660, Mexico strain) to detect in-hospital hepatitis E virus infection cases. The detecting rate of IgG was 91%, and fell to 27˜50% after 6-12 month. When he used 3-2(B) (The same ORF2 fragments from Burma strain) in detection, the rate was just 64% and no positive result was found after 6-12 months. On the contrary, 3-2(M) could not react with the convalescent serum in a Parkistan subject, but 3-2(B) could react with the serum of the same case 4.5 years later. For those proteins, when the antibody in some cases turned negative, in others it still remained in high titer. The results were similar when the mosaic protein with several linear epitopes expressed in E. Coli was used. Lack of good HEV antibody kits limited deep research on the dynamic of antibody during HIV infection. In general, during HIV infection, specific IgG antibody is detectable in early stage, peaks after 2˜4 weeks and declines quickly. Most turned to negative after 9 months, but some patients kept positive many years later. Recently, several recombinant antigens have been expressed in both baculovirus and E. coli, which reacts strongly with sera from both acute and convalescence phase. In principle, these antigens are more suitable for sero-epidemiological survey: the titer of serum HEV IgG fell rapidly after acute stage, but it still retained at a detectable level. It's worthy of notification that the antibody is related to the protection from infection during the disease epidemic.
In the other way, due to the fact that only indirect methods of detecting methods are available up to date, no established HEV IgM kit is ever developed around world. With respect to the indirect methods, on the one hand, the detect result is affected by various factors and reproduction is poor; on the other, the reliability of the result is poor for high possibility of false positive, higher negative value, or lower sensitivity. According to recent reports, during the detection of clinical samples, when the result for IgM is positive, IgG is generally positive too, thus its value on early diagnosis is limited but may be helpful in elevation of the specificity of diagnosis of acute infection.
It's reported that the antibody to several synthesized peptide and some recombinant antigens will disappear quickly in many infected subjects, so clinical diagnosis on acute hepatitis E virus infection is generally based on IgG antibody with a higher clinical concordance, but the most important defect of this method is that it can not distinguish past infection from recent infection, which will lead to both false diagnosis in high hepatitis E virus endemic area and under evaluation of the prevalence of hepatitis E virus infection during epidemiological studies. Therefore, there is urge demand to develop a reliable and sensitive anti-μ chain IgM diagnostic kit and IgG diagnostic kit which is characterized by high sensitivity toward convalescence sera.
In recent years, there was some progress in developing a highly sensitive IgG diagnostic kit. Mast et al. (1998, Hepatology, 27: 857-861) provided a comprehensive evaluation of 10 major IgG antibody EIAs around world. The concordance of many kits was fairly good in detecting known positive sera, but great difference existed among different kits in detecting American blood donors. It implied that the reliability of the results is worse in HEV prevalence studies in non-endemic areas. Among those kits, most antigens are based on linear epitopes of HEV, but two kits used conformational epitopes as antigens. First is ORF2.1 aa394˜660), the other is baculovirus expressed VLP (aa112˜607). Both antigens can detect convalescence antibody, but direct data on the comparison of the concordance between those two antigens is not available till now. It's possible that those two antigens identified different antibodies. In addition, nearly 20% prevalence was reported in non-endemic America using VLP kit, which aroused the suspicion of its specificity. But with the reported positive HIV infection in swine, goats, cows, chickens, rats, wild monkeys and enclosed monkeys, together with separately 77% and 44% antibody prevalence among wild rats in Maryland and Louisanna, it's possible that the antibody prevalence is underestimated in American population, though animal HEV can not cause clinical disease due to its virulence. (Kabrane-Lazizi et al., 1999 μm Tropic Medicine, 61:331-335). And ORF2.1 kit can detect higher positive rate among CMV infection and autoimmunological diseases. In addition, the reported ORF2.1 polypeptide, which is a GST-conchimeric protein or polyarginine conchimeric protein, intends to obtain false positive results in practice.
Both cell and tissue culture for HEV have ever been successful, and practical methods to get a large amount of virus is not yet available, so it's the only research way to switch from tradition killed or attenuated vaccine to subunit or DNA vaccine through genetic engineering.
HEV ORF2, beginning at the base positioned at 5147, has 1980 nucleotides, which encodes a polypeptide with 660 amino acids presumed to be major structural protein and constitutes the capsid of virus. At N terminal of ORF2 protein, there is a classical signal sequence followed by a region rich in arginine, which is highly positive charged region and believed to involve in genomic RNA encapisidation during virus assembly. During translation process, ORF2 entered endoplasmic reticulum (ER) by a mechanism of signal peptide recognized protein (SRP), and is further glycosylated and accumulated in ER, then probably formed the capsomer of capsid in suit. Three N-glycosylated sites, Asn-137, Asn-310 and Asn-562, are located at ORF2. They are highly conservative among different virus strains, and Asn-310 is the major glycosylated site. ORF2-transfected mammalian cell COS, human hepatocellular carcinoma Huh-7, HepG2 can thereby express a 88 kD glycoprotein which can be found in both cytoplasma and membrane. The mutation in those glycosylated sites did not affect the location of PORF2 onto cell membrane. However after the signal peptide sequence was removed therefrom, PORF2 can only be found in cytoplasma. This implied that the shift of PORF2 instead of glocosylation is necessary to protein location onto cell membrane. Like MS protein in HBV, PORF2 is possibly secreted to cell membrane directly through ER instead of Golgi body. On the surface of transfected cell, gpORF2 is not randomly distributed, but concentrated in some zone, which implied an active combination process of a protein subunit and maybe aggregate into some more ordered advance forms. The final assembly/maturation of the virus need the encapisidation of genomic RNA, thus it must be occurred in cytoplasma outside of ER or endo-wall of cell membrane. The accumulation of gpORF2 in membrane may imply the assembly of virus. At the same time, the location of capsid protein on membrane also implied the possibility of secretion of matured virus out of the cell through budding. One more attention should be drawn that, the in vitro transcript and translation of PORF2 using in-vitro translation system with translating and modifying function can produce 88 kD of gpORF2 in forms of both monomers and dimers. It illustrated that gpORF2 was prone to form homologous dimer, and capsomer of capsid may be constituted by said homologuos dimer of gp ORF2 (Jameel et al., 1996. J. Virol., 70:207-216.). Through Frost Etching election microscope, Li et al. found that recombinant HEV VLPs which is expressed by baculovirus had an icosahedral symmetry virons (T=1), which was made up of sixty p50 subunits with 22-23 nm in diameter. Since the inner space of this size particle can contain about 1 kb RNA, and HEV genome is 7.5 kb in length, it is speculated natural HEV should be a crystal lattice structure with T.gtoreq.3, but the topological structure of capsomer is similar. The total number of T=3 subunit is 90 capsomers (Li et al., 1999. virology, 265:35-45.).
According to the above, HEV is a non-enveloped virus. Virus capsid is made up of ORF2-encoded protein. The protein embodies major immunological epitopes and some neutralizing epitopes, thus it became the most favorable fragment during subunit vaccine research.
In U.S. Pat. No. 5,885,768, Reyes et al. firstly reported that 4 cynomolgus monkeys were injected i.m. with recombinant protein tipE-C2 expressed in E. Coli comprising HEV Burma strain ORF2 C terminal 2/3 (aa225˜660), wherein said protein is formulated with an alum adjuvant, by administering at 0, 30 day for 50 μg/dose. Another 2 monkeys were used as controls with adjuvant only. Four weeks later, no positive result regarding raised antibody from collected bloods is found by Western Blotting. A third-time immunization on two monkeys among them by administering 80 μg unsolvable recombinant protein without adjuvant. Four weeks later, both monkeys were positive (WB). Then the six monkeys were grouped into first and second group, each included three monkeys, two of them is immunized with either three-times or two-times inoculation, and one is control. The first group was attacked with Burma HEV, and the second group Mexico HEY. The results were, (1) ALT kept normal all the time in the immunized group, but it increased 6˜10 times higher than before immunization in control; (2) Liver biopsy sample was detected by Immunological Fluorescence method. The antigen can be found in all other monkeys except those immunized with three doses and attacked by Burma strain. (3) Virus excretion in feces can be found sequentially in all other monkeys except those immunized with three doses and attacked by Burma strain. This research sample is small, but it implied that recombinant protein from ORF2 can block the occurrence of biochemical indexes of virus hepatitis and protected completely from infection when the monkeys were attacked by wild HEV.
Tsarev et al. (1994, Proc. Nat. Acad. Sci. USA., 91:10198-10202; Tsarev et al., 1993, J. Infect. Dis., 168:369-378; Tsarev et al., 1997, Vaccine 15:1834-1838) used baculovirus in insect (SF cell) to express HEV ORF2 and got protein particles with various size from 20 nm˜30 nm in cell solution. The percentage of smaller particles is substantively increased during anaphase of infected cells. WB method was used to detect baculovirus expressed ORF2 with many specific different-size bands at 25 kD, 29 kD, 35 kD, 40-45 kD, 55˜70 kD, 72 kD. Ion exchange and molecular screening method were used to purify HEV specific protein. One day after recombinant virus infected the cells, the whole ORF2 peptide of 72 kD firstly appeared and then disappeared gradually. On the second day, the peptides of 63 kD and 55 kD appeared. On the first day, 53 kD peptide appeared in cell solution with large amount and peaked on the third day. This implied the primary 72 kD protein was randomly cut into HEV protein with 55 kD (in cell lysis solution) and/or 53 kD (cell solution). The sequencing to those two proteins showed 55 kD located at ORF2 aa112˜607, but 53 kD located at aa112˜578 and 63 kD at aa112˜660. The results of ELISA showed the activity of 55 kD was apparently stronger than 53 kD. If aa112˜660 fragment was expressed in baculovirus in insect, 63 kD and 55 kD recombinant HEV protein can also be found.
SF9 cells were collected at day 7 from the cells had been infected. The protein was primarily purified and absorbed with alum adjuvant. Then cynomolgous monkeys were immunized i.m. with 50 μg protein per dose. Four received 1 dose, the other four two doses (0 d , 28d) After the final dose, all monkeys were attacked with dose 1000˜10000CID50 i.v. of the same HEV strain (SAR-55, from a Parkistan patient). Within 15 weeks, liver biopsy, sera and feces were collected every week. Before virus attack, antibody titers in one-dose monkeys were 1:100˜1:10000, but in two-dose group they were all 1:10000 (coated with purified 55 kD). In one-dose group, one monkey was dead due to anesthesia accident 9 weeks after virus attack (still calculated in the results). In two-dose group, two monkeys died soon after the virus attack (no calculated in the results) due to unknown reason. Six monkeys after immunization were found no ALT elevation or liver biopsy pathological change, and no viremia. Among four monkeys in one-dose group, three has virus excretion, but two monkeys in two-dose group no virus excretion was found.
Further purification was done in 55 kD protein expressed in baculovirus system through ion exchange and molecular sifting methods to make its purification reachable above 99%. After absorbed with alum adjuvant, the protein with dose 50 μg, 10 μg, 2 μg, 0.4 μg, 0 μg (control) was each injected into 4 rhesus monkeys administered 0 and 28 day. Four weeks after the last dose, the monkeys were attacked with the same virus (SAR-55). Sixteen monkeys in the immunized group were all normal, and just one monkey with 2 μg dose and the other with 0.4 μg dose appeared very light pathological change. Though the immunized can prevent from hepatitis but not infection. All sixteen monkeys immunized appeared virus excretion, also viremia except one monkey with 50 μg dose and the other with 10 μg dose. And the amount of virus was limited in most monkeys, but the duration has not been shortened. Another four monkeys were immunized with 2×50 μg, and attacked with 100,000 MID50 other HEV 4 weeks after the final dose. The results were similar. All four monkeys did not show ALT elevation and pathological change, but only one monkey did not show virus excretion and viremia. The amount of virus decreased apparently, but the duration has not been shortened. According to the author's opinion, the effect of complete protection on those monkeys was worse that previous time. It's possibly attributable to the amount of virus used in attack. The amount of virus in this experiment reached 300,000, but was 1000˜10000MID50 last time. One more, the titer of antibody among groups from 0.4 μg to 50 μg has showed no difference before attack.
The staffs in Genelabs company expressed ORF2 aa112˜660 using the same baculovirus in insect and got a large amount solvable recombinant 62 kD protein. After purification, the cynomolgous monkeys were immunized and protected from the attack of virus (Mexico strain) with dose 1000CID50 (3 monkeys immunized with 20 μg did not get disease. Virus excretion was not found in two monkeys, and the amount of virus excretion decreased in one monkey). (Zhang et al., 1997, Clin Diagn Lab Immunol.; 4:423-8.)
McAtee et al., (1996, Protein Expr. Purif., 8:262-270) prepared Burma ORF2 62 kD dimer expressed in recombinant baculovirus. The dimer was dissociated into two peptides separately with 56.5 kD and 58.1 kD through HPLC-MS. Peptide mass fingerprint analysis showed the N terminal of those two peptides was the same aa112, and the C terminal is separately aa637 and aa652. And 56.5 kD protein was a very good immunogen.
Anderson group in Australia (Anderson et al., 1999.1. Virol. Methods., 81:131-142; Li et al., 1994, J Clin Microbio.) 32:2060-2066; Li et al., 1997 J. Med. Virol., 52:289-300; Li et al., 2000, J. Med. Virol., 60:379-386) used ORF2 aa394˜660 (ORF2.1) expressed in E. Coli. The product is a GST-conchimeric or poly arginine protein which can form a highly conformational convalescence epitope. This epitope can detect a very high-rate convalescence sera, but it will disappear when the fragment was extended or truncated towards N terminal. The serum at 30 weeks after rats were immunized with recombinant ORF2.1 protein was used to block the serum from convalescence patients with VLP expressed in baculovirus as the coated antigen. The blocking rate will reach 81%˜86%. Monoclonal antibody was prepared with ORF2.1 protein and two monoclonal antibody 2E2 and 4B2 which can identify ORF2.1 conformational epitopes, and five possible identifiable monoclonal antibodies were obtained. The blocking rate can reach 60% whether 2E2 or 4B2 was used to block convalescence sera with VLPs as antigen. This implied that those two monoclonal antibodies can identify the epitopes which was major components of antibody identified epitopes in convalescence sera. Different data showed that ORF2.1 had major epitope structure rather similar to VLP. The antibody to the epitopes can exist for a long time in HIV infected serum. It's probably an important protective epitope, but animal protection experiment about ORF2.1 has not been reported till now.